Project 1. Elucidating mechanisms underlying mesoderm and endoderm induction and differentiation. We devised a molecular strategy for isolating genes expressed in the yolk directly below the embryonic cells that normally form mesoderm and endoderm. This region contains essential signals for mesoderm and endoderm induction. We developed a bioinformatic method for validating our molecular strategy that will be of broad use to zebrafish researchers. We also found a master regulator transcription that acts upstream of the criticial yolk signal, working in combination with the TGFbeta ligands Ndr1 and Ndr2. Previously we identified risk factors for the zebrafish form of holoprosencephaly. Background: ndr2 mutant embryos have a more limited mesoderm and endoderm deficit than ndr1;ndr2 double-mutant embryos, namely a reduction in anterior axial mesoderm and endoderm, leading to holoprosencephaly (HPE). This phenotype is seen in only a fraction of ndr1 mutants, with many appearing wild type. At NHGRI, we examined the basis of sqt phenotypic variability and found environmental and genetic factors that influence HPE incidence. We extended our analysis to a novel sqt phenotype - bifurcation at the midline and find that the same genetic and environmental risk factors for HPE increase increase midline bifurcation, but that perturbations in WNT signaling specifically increase midline bifurcation incidence without affecting HPE. Project 2. Identifying roles of RhoGTPase signaling in gastrulation movements. Background: RhoGEF proteins are positive regulators of RhoGTPases, which have profound roles in cellular movement and morphology. To identify RhoGEFs with roles in directing the morphogenetic events of gastrulation, we performed a loss-of-function screen. We identified 48 RhoGEFs expressed during early embryogenesis and determined the loss-of-function phenotypes for 23 of these, using a non-invasive embryo holding system we designed that allows for the parallel time-lapse documention of 54 embryos. We thus identified five RhoGEFs for which two independent MOs produced the same phenotype. Three of these, homologues of ARHGEF16, Frabin and Net1, respectively, disrupted epiboly. Two others, ARHGEF10 and PLEKHG4 homologues, caused post-gastrulation defects during somitogenesis stages. We are now engaged in control experiments to confirm the specificity of these phenotypes, with a focus on the epiboly mutants. We have demonstrated the specificity of the Frabin and Net1 MOs by performing mRNA rescue experiments. Project 3. Defining roles and interactions of transcription factors in the early mesoderm and endoderm as well as transcription factors in the adjacent yolk that control key signals, nutritive and morphogenetic functions of the yolk. Background: A precise separation of newly-specified germ layer precursors in zebrafish embryos, has been untenable using conventional tools. Over previous funding years we have developed and refined a novel technique for precisely microdissecting embryonic regions of interest: FACS-assisted microdissection of photolabeled cells (FAM-P). We used this method to separate mesoderm and endoderm (mesendoderm) precursor cells from ectoderm precursor cells of pre-gastrula stage embryos, and used microarrays to assess their respective transcriptomes. Over the last 12 months, we have compared the behaviors of mesendoderm and ectoderm precursors in transplantation assays and in cell cycle assays. This has shown us that both populations retain their documented properties, and has revealed cell cycle differences between these two populations. We also performed a detailed analysis of the microarray data from the previous funding period, revealing that mesendoderm precursor cells express fewer enzyme-encoding genes and demonstrating that FAM-P purification effectively depletes extra-embryonic cells. Finally, we have investigated the requirements of several mesendoderm precursor-specific genes identified in these studies, using antisense reagents. One of these, dusp4, is required for normal head development and the differentiation of late endoderm. Our manuscript describing this work is currently in press at PNAS. We are also looking at the function of the transcription factor Six4.2. We find that the transcription factor not tail can activate six4.2 and that six4.2 can suppress a slow muscle-specific gene. We created a loss-of-function Six4.2 mutant using the emerging zinc-finger nuclease technology, and we have evidence that this mutation is lethal during juvenile development. We will investigate whether or not these juvenile mutants display and defects in muscle or movement.